Search

Machine Learning (ML) algorithms have been important tools for the extraction of useful knowledge from biological sequences, particularly in healthcare, agriculture, and the environment. However, the categorical and unstructured nature of these sequences requiring usually additional feature engineering steps, before an ML algorithm can be efficiently applied. The addition of these steps to the ML algorithm creates a processing pipeline, known as end-to-end ML. Despite the excellent results obtained by applying end-to-end ML to biotechnology problems, the performance obtained depends on the expertise of the user in the components of the pipeline. In this work, we propose an end-to-end ML-based framework called BioPrediction-RPI, which can identify implicit interactions between sequences, such as pairs of non-coding RNA and proteins, without the need for specialized expertise in end-to-end ML. This framework applies feature engineering to represent each sequence by structural and topological features. These features are divided into feature groups and used to train partial models, whose partial decisions are combined into a final decision, which, provides insights to the user by giving an interpretability report. In our experiments, the developed framework was competitive when compared with various expert-created models. We assessed BioPrediction-RPI with 12 datasets when it presented equal or better performance than all tools in 40% to 100% of cases, depending on the experiment. Finally, BioPrediction-RPI can fine-tune models based on new data and perform at the same level as ML experts, democratizing end-to-end ML and increasing its access to those working in biological sciences.

Abstract This gap in knowledge regarding ecological interactions has prompted the development of various predictive approaches. Traditionally, ecological interactions have been inferred using traits. However, the lack of trait information for numerous organisms necessitates using phylogenetic data and statistical insights from interaction matrices for prediction. Previous studies have overlooked the validation of model‐predicted interactions. This study used a novel method in predicting ecological interactions using a self‐organizing map (SOM), an unsupervised machine learning algorithm. The SOM learns from the input interaction matrix by grouping the nodes into output layers based on their interactions. Subsequently, the trained model predicts the interactions as scores. To distinguish between interactions and non‐interactions, we employed F1 score maximization, setting scores above a specific threshold as interactions and the remainder as non‐interactions. We applied this method to three unipartite metawebs and one bipartite metaweb and subsequently validated the predicted interactions using two innovative approaches: taxonomic and interaction recovery validation. Our method exhibited outstanding predictive performance, particularly for large networks. Various binary classification performance indicators, including F1 score (0.84–0.97) and accuracy (0.97–0.99), indicated high performance. Moreover, the method generated minimal predicted interactions, signifying low noise in the predictions, particularly for large networks. Taxonomic validation excels in metawebs with a connectance >0.1 but performs poorly in metawebs with very low connectance. In contrast, interaction recovery was most effective in larger metawebs. Our proposed method excels at making highly accurate predictions of ecological interactions with minimal noise, solely utilizing input interaction data without relying on traits or phylogenetic information regarding interacting nodes. These predictions are particularly precise for large networks, underscoring their potential to address knowledge gaps in emerging extensive metawebs. Notably, taxonomic validation and interaction recovery methods are sensitive to connectance and network size, respectively, suggesting prospects for developing robust interaction validation methods.

This gap in knowledge regarding ecological interactions has prompted the development of various predictive approaches. Traditionally, ecological interactions have been inferred using traits. However, the lack of trait information for numerous organisms necessitates using phylogenetic data and statistical insights from interaction matrices for prediction. Previous studies have overlooked the validation of model‐predicted interactions.This study used a novel method in predicting ecological interactions using a self‐organizing map (SOM), an unsupervised machine learning algorithm. The SOM learns from the input interaction matrix by grouping the nodes into output layers based on their interactions. Subsequently, the trained model predicts the interactions as scores. To distinguish between interactions and non‐interactions, we employed F1 score maximization, setting scores above a specific threshold as interactions and the remainder as non‐interactions. We applied this method to three unipartite metawebs and one bipartite metaweb and subsequently validated the predicted interactions using two innovative approaches: taxonomic and interaction recovery validation.Our method exhibited outstanding predictive performance, particularly for large networks. Various binary classification performance indicators, including F1 score (0.84–0.97) and accuracy (0.97–0.99), indicated high performance. Moreover, the method generated minimal predicted interactions, signifying low noise in the predictions, particularly for large networks. Taxonomic validation excels in metawebs with a connectance >0.1 but performs poorly in metawebs with very low connectance. In contrast, interaction recovery was most effective in larger metawebs.Our proposed method excels at making highly accurate predictions of ecological interactions with minimal noise, solely utilizing input interaction data without relying on traits or phylogenetic information regarding interacting nodes. These predictions are particularly precise for large networks, underscoring their potential to address knowledge gaps in emerging extensive metawebs. Notably, taxonomic validation and interaction recovery methods are sensitive to connectance and network size, respectively, suggesting prospects for developing robust interaction validation methods. [ABSTRACT FROM AUTHOR]

Drug‐drug interaction (DDI) event prediction is a challenging problem, and accurate prediction of DDI events is critical to patient health and new drug development. Recently, many machine learning‐based techniques have been proposed for predicting DDI events. However, most of the existing methods do not effectively integrate the multidimensional features of drugs and provide poor mitigation of noise to get effective feature information. To address these limitations, we propose a DDI‐Transform neural network framework for DDI event prediction. In DDI‐Transform, we design a drug structure information feature extraction module and a drug bind‐protein feature extraction module to obtain multidimensional feature information. A stack of DDI‐Transform layers in the DDI‐Transform network module are then used for adaptive learning, thus adaptively selecting the effective feature information for prediction. The results show that DDI‐Transform can accurately predict DDI events and outperform the state‐of‐the‐art models. Results on different scale datasets confirm the robustness of the method. [ABSTRACT FROM AUTHOR]

Abstract Several experimental evidences have shown that the human endogenous hormones can interact with drugs in many ways and affect drug efficacy. The hormone drug interactions (HDI) are essential for drug treatment and precision medicine; therefore, it is essential to understand the hormone-drug associations. Here, we present HormoNet to predict the HDI pairs and their risk level by integrating features derived from hormone and drug target proteins. To the best of our knowledge, this is one of the first attempts to employ deep learning approach for prediction of HDI prediction. Amino acid composition and pseudo amino acid composition were applied to represent target information using 30 physicochemical and conformational properties of the proteins. To handle the imbalance problem in the data, we applied synthetic minority over-sampling technique technique. Additionally, we constructed novel datasets for HDI prediction and the risk level of their interaction. HormoNet achieved high performance on our constructed hormone-drug benchmark datasets. The results provide insights into the understanding of the relationship between hormone and a drug, and indicate the potential benefit of reducing risk levels of interactions in designing more effective therapies for patients in drug treatments. Our benchmark datasets and the source codes for HormoNet are available in: https://github.com/EmamiNeda/HormoNet .

Several experimental evidences have shown that the human endogenous hormones can interact with drugs in many ways and affect drug efficacy. The hormone drug interactions (HDI) are essential for drug treatment and precision medicine; therefore, it is essential to understand the hormone-drug associations. Here, we present HormoNet to predict the HDI pairs and their risk level by integrating features derived from hormone and drug target proteins. To the best of our knowledge, this is one of the first attempts to employ deep learning approach for prediction of HDI prediction. Amino acid composition and pseudo amino acid composition were applied to represent target information using 30 physicochemical and conformational properties of the proteins. To handle the imbalance problem in the data, we applied synthetic minority over-sampling technique technique. Additionally, we constructed novel datasets for HDI prediction and the risk level of their interaction. HormoNet achieved high performance on our constructed hormone-drug benchmark datasets. The results provide insights into the understanding of the relationship between hormone and a drug, and indicate the potential benefit of reducing risk levels of interactions in designing more effective therapies for patients in drug treatments. Our benchmark datasets and the source codes for HormoNet are available in: https://github.com/EmamiNeda/HormoNet. [ABSTRACT FROM AUTHOR]

This special issue focuses on computational model for drug research regarding drug bioactivity prediction, drug-related interaction prediction, modelling for immunotherapy and modelling for treatment of a specific disease, as conveyed by the following six research and four review articles. Notably, these 10 papers described a wide variety of in-depth drug research from the computational perspective and may represent a snapshot of the wide research landscape. [ABSTRACT FROM AUTHOR]

Long non-coding RNA (lncRNA) are shown to be closely associated with cancer metastatic events (CME, e.g., cancer cell invasion, intravasation, extravasation, proliferation) that collaboratively accelerate malignant cancer spread and cause high mortality rate in patients. Clinical trials may accurately uncover the relationships between lncRNAs and CMEs; however, it is time-consuming and expensive. With the accumulation of data, there is an urgent need to find efficient ways to identify these relationships. Herein, a graph embedding representation-based predictor (VGEA-LCME) for exploring latent lncRNA-CME associations is introduced. In VGEA-LCME, a heterogeneous combined network is constructed by integrating similarity and linkage matrix that can maintain internal and external characteristics of networks, and a variational graph auto-encoder serves as a feature generator to represent arbitrary lncRNA and CME pair. The final robustness predicted result is obtained by ensemble classifier strategy via cross-validation. Experimental comparisons and literature verification show better remarkable performance of VGEA-LCME, although the similarities between CMEs are challenging to calculate. In addition, VGEA-LCME can further identify organ-specific CMEs. To the best of our knowledge, this is the first computational attempt to discover the potential relationships between lncRNAs and CMEs. It may provide support and new insight for guiding experimental research of metastatic cancers. The source code and data are available at https://github.com/zhuyuan-cug/VGAE-LCME. • A first attempt to discover the potential relationships between lncRNAs and CMEs computationally. • A heterogeneous combined network is constructed by matrix representation. • Ensemble algorithm is used to obtain the robustness prediction results. [ABSTRACT FROM AUTHOR]

Understanding ncRNA–protein interaction is of critical importance to unveil ncRNAs' functions. Here, we propose an integrated package LION which comprises a new method for predicting ncRNA/lncRNA–protein interaction as well as a comprehensive strategy to meet the requirement of customisable prediction. Experimental results demonstrate that our method outperforms its competitors on multiple benchmark datasets. LION can also improve the performance of some widely used tools and build adaptable models for species- and tissue-specific prediction. We expect that LION will be a powerful and efficient tool for the prediction and analysis of ncRNA/lncRNA–protein interaction. The R Package LION is available on GitHub at https://github.com/HAN-Siyu/LION/. [ABSTRACT FROM AUTHOR]

Several recent studies investigate TCR-peptide/-pMHC binding prediction using machine learning or deep learning approaches. Many of these methods achieve impressive results on test sets, which include peptide sequences that are also included in the training set. In this work, we investigate how state-ofthe-art deep learning models for TCR-peptide/-pMHC binding prediction generalize to unseen peptides. We create a dataset including positive samples from IEDB, VDJdb, McPAS-TCR, and the MIRA set, as well as negative samples from both randomization and 10X Genomics assays. We name this collection of samples TChard. We propose the hard split, a simple heuristic for training/test split, which ensures that test samples exclusively present peptides that do not belong to the training set. We investigate the effect of different training/test splitting techniques on the models’ test performance, as well as the effect of training and testing the models using mismatched negative samples generated randomly, in addition to the negative samples derived from assays. Our results show that modern deep learning methods fail to generalize to unseen peptides. We provide an explanation why this happens and verify our hypothesis on the TChard dataset. We then conclude that robust prediction of TCR recognition is still far for being solved. [ABSTRACT FROM AUTHOR]

Several recent studies investigate TCR-peptide/-pMHC binding prediction using machine learning or deep learning approaches. Many of these methods achieve impressive results on test sets, which include peptide sequences that are also included in the training set. In this work, we investigate how state-of-the-art deep learning models for TCR-peptide/-pMHC binding prediction generalize to unseen peptides. We create a dataset including positive samples from IEDB, VDJdb, McPAS-TCR, and the MIRA set, as well as negative samples from both randomization and 10X Genomics assays. We name this collection of samples TChard. We propose the hard split, a simple heuristic for training/test split, which ensures that test samples exclusively present peptides that do not belong to the training set. We investigate the effect of different training/test splitting techniques on the models’ test performance, as well as the effect of training and testing the models using mismatched negative samples generated randomly, in addition to the negative samples derived from assays. Our results show that modern deep learning methods fail to generalize to unseen peptides. We provide an explanation why this happens and verify our hypothesis on the TChard dataset. We then conclude that robust prediction of TCR recognition is still far for being solved.

Biomedical interaction networks have incredible potential to be useful in the prediction of biologically meaningful interactions, identification of network biomarkers of disease, and the discovery of putative drug targets. Recently, graph neural networks have been proposed to effectively learn representations for biomedical entities and achieved state-of-the-art results in biomedical interaction prediction. These methods only consider information from immediate neighbors but cannot learn a general mixing of features from neighbors at various distances. In this paper, we present a higher-order graph convolutional network (HOGCN)to aggregate information from the higher-order neighborhood for biomedical interaction prediction. Specifically, HOGCN collects feature representations of neighbors at various distances and learns their linear mixing to obtain informative representations of biomedical entities. Experiments on four interaction networks, including protein-protein, drug-drug, drug-target, and gene-disease interactions, show that HOGCN achieves more accurate and calibrated predictions. HOGCN performs well on noisy, sparse interaction networks when feature representations of neighbors at various distances are considered. Moreover, a set of novel interaction predictions are validated by literature-based case studies. [ABSTRACT FROM AUTHOR]

Pairwise learning corresponds to the supervised learning setting where the goal is to make predictions for pairs of objects. Prominent applications include predicting drug-target or protein-protein interactions, or customer-product preferences. In this work, we present a comprehensive review of pairwise kernels, that have been proposed for incorporating prior knowledge about the relationship between the objects. Specifically, we consider the standard, symmetric and anti-symmetric Kronecker product kernels, metric-learning, Cartesian, ranking, as well as linear, polynomial and Gaussian kernels. Recently, a O (n m + n q) time generalized vec trick algorithm, where n , m , and q denote the number of pairs, drugs and targets, was introduced for training kernel methods with the Kronecker product kernel. This was a significant improvement over previous O (n 2) training methods, since in most real-world applications m , q < < n . In this work we show how all the reviewed kernels can be expressed as sums of Kronecker products, allowing the use of generalized vec trick for speeding up their computation. In the experiments, we demonstrate how the introduced approach allows scaling pairwise kernels to much larger data sets than previously feasible, and provide an extensive comparison of the kernels on a number of biological interaction prediction tasks. [ABSTRACT FROM AUTHOR]

Identifying interactions between compound and protein is a substantial part of the drug discovery process. Accurate prediction of interaction relationships can greatly reduce the time of drug development. The uniqueness of our method lies in three aspects:1) it represents a compound with a distance matrix. A distance matrix can capture the structural information, compared with the SMILES string. On the other hand, a distance matrix does not require complex data preprocessing for the molecular structure as the molecular graph representation, and is easier to obtain; 2) it uses SPP(Spatial pyramid pooling)-net to extract compound features, which has been successfully applied in image classification; and 3) it extracts protein features through the natural language processing method (doc2vec) to obtain sequence semantic information. We evaluated our method on three benchmark datasets—human, C.elegans, and DUDE—and the experimental results demonstrate that our proposed model presents competitive performance against state-of-the-art predictors. We also carried out drug–drug interaction (DDI) experiments to verify the strong potential of distance matrix as molecular characteristics. The source code and datasets are available at https://github.com/lxlsu/SPP_CPI. [ABSTRACT FROM AUTHOR]

Abstract Background Due to the lack of enough interaction data among compositions, targets and diseases, it is difficult to construct a complete network of Traditional Chinese Medicine (TCM) that comprehensively reflects active compositions and their synergistic network in terms of specific diseases. Therefore, mapping of the full spectrum of interaction between compounds and their targets is of central importance when we use network pharmacology approach to explore the therapeutic potential of the TCM. Methods To address this challenge, we developed a large-scale simultaneous interaction prediction approach (SiPA) integrated one interaction network based simple inference model (SIM), focusing on ‘logical relevance’ between compounds, proteins or diseases, and another compound-target correlation space based interaction prediction model (CTCS-IPM) that was built on the basis of the canonical correlation analysis (CCA) to estimate the position of compounds (or targets) in compound-protein correlated space. Then SiPA was applied to discover reliable multiple interactions for interaction network expansion of a TCM, compound Salvia miltiorrhiza. By means of network analysis, potential active compounds and their related network synergy underlying cardiovascular diseases were evaluated between expanded and original interaction networks. Part of new interactions were validated with existing experimental evidence and molecular docking. Results As evaluated with known test dataset, the established combination approach was proved to make highly accurate prediction, showing a well prediction performance for the SIM and a high recall rate of 85.2% for the CTCS-IPM. Then 710 pairs of new compound-target interactions, 24 pairs of new compound-cardiovascular disease interactions and 294 pairs of new cardiovascular disease-protein interactions were predicted for compound Salvia miltiorrhiza. Results of network analysis suggested the network expansion could dramatically improve the completeness and effectiveness of the network. Validation results of literature and molecular docking manifested that inferred interactions had good reliability. Conclusions We provided a practical and efficient way for large-scale inference of multiple interactions of TCM ingredients, which was not limited by the lack of negative samples, sample size and target 3D structures. SiPA could help researchers more accurately prioritize the effective compounds and more completely explore network synergy of TCM for treating specific diseases, indicating a potential way for effectively identifying candidate compound (or target) in drug discovery.

More than ten million deaths make influenza virus one of the deadliest of history. About half a million sever illnesses are annually reported consequent of influenza. Influenza is a parasite which needs the host cellular machinery to replicate its genome. To reach the host, viral proteins need to interact with the host proteins. Therefore, identification of host-virus protein interaction network (HVIN) is one of the crucial steps in treating viral diseases. Being expensive, time-consuming and laborious of HVIN experimental identification, force the researches to use computational methods instead of experimental ones to obtain a better understanding of HVIN. In this study, several features are extracted from physicochemical properties of amino acids, combined with different centralities of human protein-protein interaction network (HPPIN) to predict protein-protein interactions between human proteins and Alphainfluenzavirus proteins (HI-PPIs). Ensemble learning methods were used to predict such PPIs. Our model reached 0.93 accuracy, 0.91 sensitivity and 0.95 specificity. Moreover, a database including 694522 new PPIs was constructed by prediction results of the model. Further analysis showed that HPPIN centralities, gene ontology semantic similarity and conjoint triad of virus proteins are the most important features to predict HI-PPIs.

Abstract Background Computational prediction of drug-target interactions (DTI) is vital for drug discovery. The experimental identification of interactions between drugs and target proteins is very onerous. Modern technologies have mitigated the problem, leveraging the development of new drugs. However, drug development remains extremely expensive and time consuming. Therefore, in silico DTI predictions based on machine learning can alleviate the burdensome task of drug development. Many machine learning approaches have been proposed over the years for DTI prediction. Nevertheless, prediction accuracy and efficiency are persisting problems that still need to be tackled. Here, we propose a new learning method which addresses DTI prediction as a multi-output prediction task by learning ensembles of multi-output bi-clustering trees (eBICT) on reconstructed networks. In our setting, the nodes of a DTI network (drugs and proteins) are represented by features (background information). The interactions between the nodes of a DTI network are modeled as an interaction matrix and compose the output space in our problem. The proposed approach integrates background information from both drug and target protein spaces into the same global network framework. Results We performed an empirical evaluation, comparing the proposed approach to state of the art DTI prediction methods and demonstrated the effectiveness of the proposed approach in different prediction settings. For evaluation purposes, we used several benchmark datasets that represent drug-protein networks. We show that output space reconstruction can boost the predictive performance of tree-ensemble learning methods, yielding more accurate DTI predictions. Conclusions We proposed a new DTI prediction method where bi-clustering trees are built on reconstructed networks. Building tree-ensemble learning models with output space reconstruction leads to superior prediction results, while preserving the advantages of tree-ensembles, such as scalability, interpretability and inductive setting.

Various studies on latency in remote mixed reality collaborations (remote MR collaboration) have been conducted, but studies related to interaction latency are scarce. Interaction latency in a remote MR collaboration occurs because action detection (such as contact or collision) between a human and a virtual object is required for finding the interaction performed. Therefore, in this paper, we propose a method based on interaction prediction to reduce the time for detecting the action between humans and virtual objects. The proposed method predicts an interaction based on consecutive joint angles. To examine the effectiveness of the proposed method, an experiment was conducted and the results were given. From the experimental results, it was confirmed that the proposed method could reduce the interaction latency compared to the one obtained by conventional methods. [ABSTRACT FROM AUTHOR]

LncRNA–miRNA interactions contribute to the regulation of therapeutic targets and diagnostic biomarkers in multifarious human diseases. However, it remains difficult to experimentally identify lncRNA–miRNA associations at large scale, and computational prediction methods are limited. In this study, we developed a network distance analysis model for lncRNA–miRNA association prediction (NDALMA). Similarity networks for lncRNAs and miRNAs were calculated and integrated with Gaussian interaction profile (GIP) kernel similarity. Then, network distance analysis was applied to the integrated similarity networks, and final scores were obtained after confidence calculation and score conversion. Our model obtained satisfactory results in fivefold cross validation, achieving an AUC of 0.8810 and an AUPR of 0.8315. Moreover, NDALMA showed superior prediction performance over several other network algorithms, and we tested the suitability and flexibility of the model by comparing different types of similarity. In addition, case studies of the relationships between lncRNAs and miRNAs were conducted, which verified the reliability of our method in predicting lncRNA–miRNA associations. The datasets and source code used in this study are available at https://github.com/Liu-Lab-Lnu/NDALMA. [ABSTRACT FROM AUTHOR]

Abstract Background Network inference is crucial for biomedicine and systems biology. Biological entities and their associations are often modeled as interaction networks. Examples include drug protein interaction or gene regulatory networks. Studying and elucidating such networks can lead to the comprehension of complex biological processes. However, usually we have only partial knowledge of those networks and the experimental identification of all the existing associations between biological entities is very time consuming and particularly expensive. Many computational approaches have been proposed over the years for network inference, nonetheless, efficiency and accuracy are still persisting open problems. Here, we propose bi-clustering tree ensembles as a new machine learning method for network inference, extending the traditional tree-ensemble models to the global network setting. The proposed approach addresses the network inference problem as a multi-label classification task. More specifically, the nodes of a network (e.g., drugs or proteins in a drug-protein interaction network) are modelled as samples described by features (e.g., chemical structure similarities or protein sequence similarities). The labels in our setting represent the presence or absence of links connecting the nodes of the interaction network (e.g., drug-protein interactions in a drug-protein interaction network). Results We extended traditional tree-ensemble methods, such as extremely randomized trees (ERT) and random forests (RF) to ensembles of bi-clustering trees, integrating background information from both node sets of a heterogeneous network into the same learning framework. We performed an empirical evaluation, comparing the proposed approach to currently used tree-ensemble based approaches as well as other approaches from the literature. We demonstrated the effectiveness of our approach in different interaction prediction (network inference) settings. For evaluation purposes, we used several benchmark datasets that represent drug-protein and gene regulatory networks. We also applied our proposed method to two versions of a chemical-protein association network extracted from the STITCH database, demonstrating the potential of our model in predicting non-reported interactions. Conclusions Bi-clustering trees outperform existing tree-based strategies as well as machine learning methods based on other algorithms. Since our approach is based on tree-ensembles it inherits the advantages of tree-ensemble learning, such as handling of missing values, scalability and interpretability.

In the development and treatment of many human diseases, the regulatory roles between lncRNAs and miRNAs are important, but much remains unknown about them; moreover, experimental methods for analyzing them are expensive and time-consuming. In this work, we applied a semi-supervised interactome network-based approach to explore and forecast the latent interaction between lncRNAs and miRNAs. We constructed graphs according to the similarity of each of lncRNAs and miRNAs and determined the number of graphlet interaction isomers between nodes in these two graphs. According to the two graphs and the known interactive relationship, we calculated a score for lncRNA–miRNA pairs, as the prediction result. The results showed that the model (LMI-INGI) was reliable in fivefold cross-validation (AUC = 0.8957, PRE = 0.6815, REC = 0.8842, F1 score = 0.7452, AUPR = 0.9213). We also tested the model with data based on the similarity of expression profile and similarity of function for verifying the applicability of LMI-INGI, and the resulting AUC value was 0.9197 and 0.9006, respectively. Compared with the other four algorithms and variable similarity tests, our model successfully demonstrated superiority and good generalizability. LMI-INGI would be helpful in forecasting interactions between lncRNAs and miRNAs. • The computational model made use of semi-supervised interactome network-based approach. • The AUC of 5-fold cross validation (5-fold CV) were significantly better than other previous computational methods. • The similarity of expression profile and similarity of function for verifying the applicability of LMI-INGI were tested. • LMI-INGI would be helpful in forecasting interactions between lncRNAs and miRNAs. [ABSTRACT FROM AUTHOR]

The recent emergence of deep learning to characterize complex patterns of protein big data reveals its potential to address the classic challenges in the field of protein data mining. Much research has revealed the promise of deep learning as a powerful tool to transform protein big data into valuable knowledge, leading to scientific discoveries and practical solutions. In this review, we summarize recent publications on deep learning predictive approaches in the field of mining protein data. The application architectures of these methods include multilayer perceptrons, stacked autoencoders, deep belief networks, two- or three-dimensional convolutional neural networks, recurrent neural networks, graph neural networks, and complex neural networks and are described from five perspectives: residue-level prediction, sequence-level prediction, three-dimensional structural analysis, interaction prediction, and mass spectrometry data mining. The advantages and deficiencies of these architectures are presented in relation to various tasks in protein data mining. Additionally, some practical issues and their future directions are discussed, such as robust deep learning for protein noisy data, architecture optimization for specific tasks, efficient deep learning for limited protein data, multimodal deep learning for heterogeneous protein data, and interpretable deep learning for protein understanding. This review provides comprehensive perspectives on general deep learning techniques for protein data analysis. [ABSTRACT FROM AUTHOR]

Machine Learning (ML) algorithms have been important tools for the extraction of useful knowledge from biological sequences, particularly in healthcare, agriculture, and the environment. However, the categorical and unstructured nature of these sequences requiring usually additional feature engineering steps, before an ML algorithm can be efficiently applied. The addition of these steps to the ML algorithm creates a processing pipeline, known as end-to-end ML. Despite the excellent results obtained by applying end-to-end ML to biotechnology problems, the performance obtained depends on the expertise of the user in the components of the pipeline. In this work, we propose an end-to-end ML-based framework called BioPrediction-RPI, which can identify implicit interactions between sequences, such as pairs of non-coding RNA and proteins, without the need for specialized expertise in end-to-end ML. This framework applies feature engineering to represent each sequence by structural and topological features. These features are divided into feature groups and used to train partial models, whose partial decisions are combined into a final decision, which, provides insights to the user by giving an interpretability report. In our experiments, the developed framework was competitive when compared with various expert-created models. We assessed BioPrediction-RPI with 12 datasets when it presented equal or better performance than all tools in 40% to 100% of cases, depending on the experiment. Finally, BioPrediction-RPI can fine-tune models based on new data and perform at the same level as ML experts, democratizing end-to-end ML and increasing its access to those working in biological sciences., Competing Interests: We have no conflicts of interest to declare., (© 2024 The Authors.)

Abstract Background Protein-protein interaction (PPI) prediction remains a central task in systems biology to achieve a better and holistic understanding of cellular and intracellular processes. Recently, an increasing number of computational methods have shifted from pair-wise prediction to network level prediction. Many of the existing network level methods predict PPIs under the assumption that the training network should be connected. However, this assumption greatly affects the prediction power and limits the application area because the current golden standard PPI networks are usually very sparse and disconnected. Therefore, how to effectively predict PPIs based on a training network that is sparse and disconnected remains a challenge. Results In this work, we developed a novel PPI prediction method based on deep learning neural network and regularized Laplacian kernel. We use a neural network with an autoencoder-like architecture to implicitly simulate the evolutionary processes of a PPI network. Neurons of the output layer correspond to proteins and are labeled with values (1 for interaction and 0 for otherwise) from the adjacency matrix of a sparse disconnected training PPI network. Unlike autoencoder, neurons at the input layer are given all zero input, reflecting an assumption of no a priori knowledge about PPIs, and hidden layers of smaller sizes mimic ancient interactome at different times during evolution. After the training step, an evolved PPI network whose rows are outputs of the neural network can be obtained. We then predict PPIs by applying the regularized Laplacian kernel to the transition matrix that is built upon the evolved PPI network. The results from cross-validation experiments show that the PPI prediction accuracies for yeast data and human data measured as AUC are increased by up to 8.4 and 14.9% respectively, as compared to the baseline. Moreover, the evolved PPI network can also help us leverage complementary information from the disconnected training network and multiple heterogeneous data sources. Tested by the yeast data with six heterogeneous feature kernels, the results show our method can further improve the prediction performance by up to 2%, which is very close to an upper bound that is obtained by an Approximate Bayesian Computation based sampling method. Conclusions The proposed evolution deep neural network, coupled with regularized Laplacian kernel, is an effective tool in completing sparse and disconnected PPI networks and in facilitating integration of heterogeneous data sources.

Proteins are the most fascinating multifaceted biomacromolecules in living systems and play various important roles such as structural, sensory, catalytic, and regulatory function. Protein and peptide interactions have emerged as an important and challenging topic in biochemistry and medicinal chemistry. Computational methods as promising tools have been utilized to predict protein and peptide interactions in order to intervene in the biochemical processes and facilitate pharmaceutical peptide design and clarify the complications. This review will introduce the computational methods which are applicable in protein and peptide interaction prediction and summarizes the most successful examples of computational methods described in the literature. HIGHLIGHTS •Highlights the importance of peptides and proteins interactions. •Summarizes the computational methods which are applicable in peptide and protein interaction prediction. •Highlights the applications of computational methods in peptides and proteins interactions.

Predicting the interaction between two humans, when viewed as a part of video is one of the most challenging issues in the field of computer vision, due to its various applications. This paper presents a new interaction prediction method that has a high accuracy in detecting the interactions when a small percentage of the video is viewed. At first, the interacting people are detected and then a dual-actor CNN model is utilized to recognize the type of interaction between the detected people. This model consists of two CNN networks while the parameters of which are shared. Each branch of this model extracts deep temporal or spatial features. The spatial and the temporal models are learned with Long Short Term Memory (LSTM) networks to model time information. Finally, the spatial and temporal models are combined to predict the interaction. The results show that the proposed model gives improvements on standard interaction recognition datasets including the TV Human Interaction, BIT interaction and UT Interaction. [ABSTRACT FROM AUTHOR]

Recently, the prediction of interactions in videos has been an active subject in computer vision. Its goal is to deduce interactions in their early stages. Many approaches have been proposed to predict interaction, but it still remains a challenging problem. In the present paper, features are optical flow fields extracted from video frames using convolutional neural networks. This feature, which is extracted from successive frames, constructs a time series. Then, the problem is modeled in the form of a time series prediction. Prediction of the interaction type is based on matching the time series under experiment with the time series available in the training set. Dynamic time warping provides an optimal match between a pair of time-series data by a nonlinear mapping between two data. Finally, the SVM and KNN classification methods with dynamic time warping distance are used to predict the video label. The results showed that the proposed model improved on standard interaction recognition datasets including the TVHI, BIT, and UT interaction. [ABSTRACT FROM AUTHOR]

Recently, lncRNAs have attracted accumulating attentions because more and more experimental researches have shown lncRNA can play critical roles in many biological processes. Predicting potential interactions between lncRNAs and proteins are key to understand the lncRNAs biological functions. But traditional biological experiments are expensive and time-consuming, network similarity methods provide a powerful solution to computationally predict lncRNA-protein interactions. In this work, a novel path-based lncRNA-protein interaction (PBLPI) prediction model is proposed by integrating protein semantic similarity, lncRNA functional similarity, known human lncRNA-protein interactions, and Gaussian interaction profile kernel similarity. PBLPI model utilizes three interlinked sub-graphs to construct a heterogeneous graph, and then infers potential lncRNA-protein interactions through depth-first search algorithm. Consequently, PBLPI achieves reliable performance in the frameworks of 5-fold cross validation (average AUC is 0.9244 and AUPR is 0.6478). In the case study, we use " Mus musculus " data to further validate the reliability of PBLPI method. It is anticipated that PBLPI would become a useful tool to identify potential lncRNA-protein interactions. • The computational model made use of heterogeneous graph and depth-first search algorithm. • The AUC of 5-fold cross validation (5-fold CV) were significantly better than other previous computational methods. • The case study could further reflect the reliability of PBLPI • PBLPI could be a reliable method for proteins associated lncRNAs prediction. [ABSTRACT FROM AUTHOR]

Background: Computational prediction of drug-target interactions (DTI) is vital for drug discovery. The experimental identification of interactions between drugs and target proteins is very onerous. Modern technologies have mitigated the problem, leveraging the development of new drugs. However, drug development remains extremely expensive and time consuming. Therefore, in silico DTI predictions based on machine learning can alleviate the burdensome task of drug development. Many machine learning approaches have been proposed over the years for DTI prediction. Nevertheless, prediction accuracy and efficiency are persisting problems that still need to be tackled. Here, we propose a new learning method which addresses DTI prediction as a multi-output prediction task by learning ensembles of multi-output bi-clustering trees (eBICT) on reconstructed networks. In our setting, the nodes of a DTI network (drugs and proteins) are represented by features (background information). The interactions between the nodes of a DTI network are modeled as an interaction matrix and compose the output space in our problem. The proposed approach integrates background information from both drug and target protein spaces into the same global network framework. Results: We performed an empirical evaluation, comparing the proposed approach to state of the art DTI prediction methods and demonstrated the effectiveness of the proposed approach in different prediction settings. For evaluation purposes, we used several benchmark datasets that represent drug-protein networks. We show that output space reconstruction can boost the predictive performance of tree-ensemble learning methods, yielding more accurate DTI predictions. Conclusions: We proposed a new DTI prediction method where bi-clustering trees are built on reconstructed networks. Building tree-ensemble learning models with output space reconstruction leads to superior prediction results, while preserving the advantages of tree-ensembles, such as scalability, interpretability and inductive setting. [ABSTRACT FROM AUTHOR]

Various studies on latency in remote mixed reality collaborations (remote MR collaboration) have been conducted, but studies related to interaction latency are scarce. Interaction latency in a remote MR collaboration occurs because action detection (such as contact or collision) between a human and a virtual object is required for finding the interaction performed. Therefore, in this paper, we propose a method based on interaction prediction to reduce the time for detecting the action between humans and virtual objects. The proposed method predicts an interaction based on consecutive joint angles. To examine the effectiveness of the proposed method, an experiment was conducted and the results were given. From the experimental results, it was confirmed that the proposed method could reduce the interaction latency compared to the one obtained by conventional methods.

According to the latest research, lncRNAs (long non-coding RNAs) play a broad and important role in various biological processes by interacting with proteins. However, identifying whether proteins interact with a specific lncRNA through biological experimental methods is difficult, costly, and time-consuming. Thus, many bioinformatics computational methods have been proposed to predict lncRNA-protein interactions. In this paper, we proposed a novel approach called Long non-coding RNA-Protein Interaction Prediction based on Improved Bipartite Network Recommender Algorithm (LPI-IBNRA). In the proposed method, we implemented a two-round resource allocation and eliminated the second-order correlations appropriately on the bipartite network. Experimental results illustrate that LPI-IBNRA outperforms five previous methods, with the AUC values of 0.8932 in leave-one-out cross validation (LOOCV) and 0.8819 ± 0.0052 in 10-fold cross validation, respectively. In addition, case studies on four lncRNAs were carried out to show the predictive power of LPI-IBNRA.

Antibodies are a class of proteins that recognize and neutralize pathogens by binding to their antigens. They are the most significant category of biopharmaceuticals for both diagnostic and therapeutic applications. Understanding how antibodies interact with their antigens plays a fundamental role in drug and vaccine design and helps to comprise the complex antigen binding mechanisms. Computational methods for predicting interaction sites of antibody-antigen are of great value due to the overall cost of experimental methods. Machine learning methods and deep learning techniques obtained promising results.In this work, we predict antibody interaction interface sites by applying HSS-PPI, a hybrid method defined to predict the interface sites of general proteins. The approach abstracts the proteins in terms of hierarchical representation and uses a graph convolutional network to classify the amino acids between interface and non-interface. Moreover, we also equipped the amino acids with different sets of physicochemical features together with structural ones to describe the residues. Analyzing the results, we observe that the structural features play a fundamental role in the amino acid descriptions. We compare the obtained performances, evaluated using standard metrics, with the ones obtained with SVM with 3D Zernike descriptors, Parapred, Paratome, and Antibody i-Patch., (© 2024. The Author(s), under exclusive license to Springer Science+Business Media, LLC, part of Springer Nature.)

Background: Network inference is crucial for biomedicine and systems biology. Biological entities and their associations are often modeled as interaction networks. Examples include drug protein interaction or gene regulatory networks. Studying and elucidating such networks can lead to the comprehension of complex biological processes. However, usually we have only partial knowledge of those networks and the experimental identification of all the existing associations between biological entities is very time consuming and particularly expensive. Many computational approaches have been proposed over the years for network inference, nonetheless, efficiency and accuracy are still persisting open problems. Here, we propose bi-clustering tree ensembles as a new machine learning method for network inference, extending the traditional tree-ensemble models to the global network setting. The proposed approach addresses the network inference problem as a multi-label classification task. More specifically, the nodes of a network (e.g., drugs or proteins in a drug-protein interaction network) are modelled as samples described by features (e.g., chemical structure similarities or protein sequence similarities). The labels in our setting represent the presence or absence of links connecting the nodes of the interaction network (e.g., drug-protein interactions in a drug-protein interaction network). Results: We extended traditional tree-ensemble methods, such as extremely randomized trees (ERT) and random forests (RF) to ensembles of bi-clustering trees, integrating background information from both node sets of a heterogeneous network into the same learning framework. We performed an empirical evaluation, comparing the proposed approach to currently used tree-ensemble based approaches as well as other approaches from the literature. We demonstrated the effectiveness of our approach in different interaction prediction (network inference) settings. For evaluation purposes, we used several benchmark datasets that represent drug-protein and gene regulatory networks. We also applied our proposed method to two versions of a chemical-protein association network extracted from the STITCH database, demonstrating the potential of our model in predicting non-reported interactions. Conclusions: Bi-clustering trees outperform existing tree-based strategies as well as machine learning methods based on other algorithms. Since our approach is based on tree-ensembles it inherits the advantages of tree-ensemble learning, such as handling of missing values, scalability and interpretability. [ABSTRACT FROM AUTHOR]

According to the latest research, lncRNAs (long non-coding RNAs) play a broad and important role in various biological processes by interacting with proteins. However, identifying whether proteins interact with a specific lncRNA through biological experimental methods is difficult, costly, and time-consuming. Thus, many bioinformatics computational methods have been proposed to predict lncRNA-protein interactions. In this paper, we proposed a novel approach called Long non-coding RNA-Protein Interaction Prediction based on Improved Bipartite Network Recommender Algorithm (LPI-IBNRA). In the proposed method, we implemented a two-round resource allocation and eliminated the second-order correlations appropriately on the bipartite network. Experimental results illustrate that LPI-IBNRA outperforms five previous methods, with the AUC values of 0.8932 in leave-one-out cross validation (LOOCV) and 0.8819 ± 0.0052 in 10-fold cross validation, respectively. In addition, case studies on four lncRNAs were carried out to show the predictive power of LPI-IBNRA. [ABSTRACT FROM AUTHOR]

Long non-coding RNA (lncRNA) plays an important role in many important biological processes and has attracted widespread attention. Although the precise functions and mechanisms for most lncRNAs are still unknown, we are certain that lncRNAs usually perform their functions by interacting with the corresponding RNA- binding proteins. For example, lncRNA-protein interactions play an important role in post transcriptional gene regulation, such as splicing, translation, signaling, and advances in complex diseases. However, experimental verification of lncRNA-protein interactions prediction is time-consuming and laborious. In this work, we propose a computational method, named IRWNRLPI, to find the potential associations between lncRNAs and proteins. IRWNRLPI integrates two algorithms, random walk and neighborhood regularized logistic matrix factorization, which can optimize a lot more than using an algorithm alone. Moreover, the method is semi-supervised and does not require negative samples. Based on the leave-one-out cross validation, we obtain the AUC of 0.9150 and the AUPR of 0.7138, demonstrating its reliable performance. In addition, by means of case study in the “Mus musculus,” many lncRNA-protein interactions which are predicted by our method can be successfully confirmed by experiments. This suggests that IRWNRLPI will be a useful bioinformatics resource in biomedical research.

Tea leaf spot caused by Didymella bellidis can seriously reduce the productivity and quality of tea ( Camellia sinensis var. sinensis ) leaves in Guizhou Province, southwest China. Analysis of the relationship between messenger RNAs (mRNAs) and long non-coding RNAs (lncRNAs) of tea could provide insights into the plant-pathogen interaction. In this study, high-throughput sequencing of mRNAs and lncRNAs from tea leaves during infection by D. bellidis was conducted using the Illumina Novaseq 6000 platform. Infection by D. bellidis hyphae resulted in up- or downregulation of 553 and 191 of the differentially expressed mRNAs (DEmRNAs), respectively. As the S gene number (total number of genes with significantly differential expression annotated in the specified Gene Ontology [GO] database), three were enriched with respect to the defense response to the fungus at the biological process level. Expression of the DEmRNAs peroxidase 21 (TEA000222.1) and mcht-2 (TEA013240.1) originating from tea leaves were upregulated during challenge by D. bellidis hyphae, whereas expression of the LRR receptor-like serine/threonine-protein kinase ERECTA (TEA016781.1) gene was downregulated. The infection of D. bellidis hyphae resulted in up- or downregulation of 227 and 958 of the differentially expressed lncRNAs (DElncRNAs). The DEmRNAs associated with uncharacterized LOC101499401 (TEA015626.1), uncharacterized protein (TEA014125.1), structural maintenance of chromosomes protein 1 (TEA001660.1), and uncharacterized protein (TEA017727.1) occurred as a result of cis regulation by DElncRNAs MSTRG.20036, MSTRG.3843, MSTRG.26132, and MSTRG.56701, respectively. The expression profiling and lncRNA/mRNA association prediction in the tea leaves infected by D. bellidis will provide a valuable resource for further research into disease resistance., Competing Interests: The author(s) declare no conflict of interest.

Long non-coding RNA (lncRNA) plays an important role in many important biological processes and has attracted widespread attention. Although the precise functions and mechanisms for most lncRNAs are still unknown, we are certain that lncRNAs usually perform their functions by interacting with the corresponding RNA- binding proteins. For example, lncRNA-protein interactions play an important role in post transcriptional gene regulation, such as splicing, translation, signaling, and advances in complex diseases. However, experimental verification of lncRNA-protein interactions prediction is time-consuming and laborious. In this work, we propose a computational method, named IRWNRLPI, to find the potential associations between lncRNAs and proteins. IRWNRLPI integrates two algorithms, random walk and neighborhood regularized logistic matrix factorization, which can optimize a lot more than using an algorithm alone. Moreover, the method is semi-supervised and does not require negative samples. Based on the leave-one-out cross validation, we obtain the AUC of 0.9150 and the AUPR of 0.7138, demonstrating its reliable performance. In addition, by means of case study in the “Mus musculus,” many lncRNA-protein interactions which are predicted by our method can be successfully confirmed by experiments. This suggests that IRWNRLPI will be a useful bioinformatics resource in biomedical research. [ABSTRACT FROM AUTHOR]

Pathogens infect host organisms by exploiting host cellular mechanisms and evading host defence mechanisms through molecular pathogen–host interactions (PHIs). Discovering new interactions between pathogen and human proteins is very crucial in understanding the infection mechanisms. By analysing interaction networks, the interactions responsible for infectious diseases can be detected and new drugs disabling these interactions can be delivered. In this paper, we propose a method based on Bayesian matrix factorization for predicting PHIs along with a projection-based technique and combine the results by employing an ensemble method. Furthermore, two features, target similarity and attacker similarity , are utilized for the first time in the literature for PHI prediction. The advantages of the proposed methods are two folds. Firstly, they relieve the need for negative samples which is significant since there is no available dataset providing negative samples for most of the pathogenic systems. Secondly, the experiments demonstrate that the proposed approach outperforms state-of-the-art methods; roughly 20% of top 50 predictions are among recently validated interactions. So, the search space for wet-lab experiments to obtain validated interactions can be considerably narrowed down from a huge number of possible interactions. [ABSTRACT FROM AUTHOR]

Predicting an interaction before it is fully executed is very important in applications, such as human-robot interaction and video surveillance. In a two-human interaction scenario, there are often contextual dependency structures between the global interaction context of the two humans and the local context of the different body parts of each human. In this paper, we propose to learn the structure of the interaction contexts and combine it with the spatial and temporal information of a video sequence to better predict the interaction class. The structural models, including the spatial and the temporal models, are learned with long short term memory (LSTM) networks to capture the dependency of the global and local contexts of each RGB frame and each optical flow image, respectively. LSTM networks are also capable of detecting the key information from global and local interaction contexts. Moreover, to effectively combine the structural models with the spatial and temporal models for interaction prediction, a ranking score fusion method is introduced to automatically compute the optimal weight of each model for score fusion. Experimental results on the BIT-Interaction Dataset and the UT-Interaction Dataset clearly demonstrate the benefits of the proposed method. [ABSTRACT FROM AUTHOR]

Recent advances in Artificial Intelligence (AI) and Machine Learning (ML) brought light on the need for explainability in multiple domains (e.g., healthcare, finance, justice, and recruiting). Explainability or Explainable AI (XAI) can be defined as everything that makes AI more understandable to human beings. However, XAI features may vary according to the AI algorithm used. Beyond XAI features, different AI algorithms vary in terms of speed, performance, and costs associated with training/running models. Knowing when to choose the right algorithm for the task at hand, therefore, is fundamental in multiple AI systems, for instance, AutoML and AutoAI. In this paper, we propose a method to analyze patterns of high-granularity user interface (UI) events (i.e., mouse, keyboard, and additional custom events triggered on the millisecond scale) to predict when users will interact with UI elements that provide explainability for the AI in place. In this context, this paper presents: (1) a user study involving 37 participants (7 in the pilot phase and 30 in the main experiment phase) in which people performed a task of reporting a bug using a text form associated with an AI data quality meter and its XAI UI element and (2) an approach to model micro behavior using n o d e 2 v e c to predict when the interaction with XAI UI element will occur. The proposed approach uses a rich dataset (approximately 129k events) and combines n o d e 2 v e c and a Logistic Regression classifier. Results obtained show we have obtained an event-by-event prediction of the interaction with XAI with an average F-score of 0.90 (σ = 0. 06). From the presented results, one expects to support researchers in the realm of UI personalization to consider high-granularity interaction data when predicting the need for XAI while users are interacting with AI model outputs. • A study detailing different behaviors while users interact with explainability. • A user model approach to predict interaction with explainability. • The proposed approach obtained an average F-score of 0.90. • Groups do not differ in terms of task time nor event stream length. • An extension for XAI taxonomy encompassing interaction with XAI. [ABSTRACT FROM AUTHOR]

Therapeutic activities of drugs are often influenced by co-administration of drugs that may cause inevitable drug-drug interactions (DDIs) and inadvertent side effects. Prediction and identification of DDIs are extremely vital for the patient safety and success of treatment modalities. A number of computational methods have been employed for the prediction of DDIs based on drugs structures and/or functions. Here, we report on a computational method for DDIs prediction based on functional similarity of drugs. The model was set based on key biological elements including carriers, transporters, enzymes and targets (CTET). The model was applied for 2189 approved drugs. For each drug, all the associated CTETs were collected, and the corresponding binary vectors were constructed to determine the DDIs. Various similarity measures were conducted to detect DDIs. Of the examined similarity methods, the inner product-based similarity measures (IPSMs) were found to provide improved prediction values. Altogether, 2,394,766 potential drug pairs interactions were studied. The model was able to predict over 250,000 unknown potential DDIs. Upon our findings, we propose the current method as a robust, yet simple and fast, universal in silico approach for identification of DDIs. We envision that this proposed method can be used as a practical technique for the detection of possible DDIs based on the functional similarities of drugs. [ABSTRACT FROM AUTHOR]